Publication Radar · Source metadata
ResCom: Activation Residual Compensation for Large Language Model Quantization
Publication record
Bibliographic evidence describes where and when a record appeared. It does not validate scientific claims.
First announced by arXiv: 2026-10-28 · Indexed by Pub-AI: 2026-10-28 UTC
ResCom: Activation Residual Compensation for Large Language Model Quantization
First announced by arXiv: 2026-10-28 · Original record ↗
- arxiv
- 2610.35955
Explicit source versions (1)
- v1 · Date not supplied
- Preprint available
Source date: 2026-10-28 · First detected: 2026-10-28 UTC · Historical baseline
Source evidence (1)
- Official evidence ↗ · RSS guid=oai:arXiv.org:2610.35955v1; arxiv:announce_type=new; item pubDate (official metadata)
Abstract · arxiv
Low-bit weight--activation quantization reduces the memory and computation needed to deploy large language models, but activation rounding can still degrade accuracy after calibration. Reconstruction from rounded inputs cannot distinguish activations within the same quantization bin. We propose ResCom, a low-rank correction that uses the activation rounding residual to compensate for the resulting layer-output error. An output-aware calibration procedure selects residual directions by their contribution to output reconstruction and computes the correction in closed form, without gradient training or updates to the calibrated weights. On Llama-3-8B with 4-bit weights and activations, rank-32 attention corrections reduce WikiText-2 perplexity from 7.708 to 7.147, improve MMLU by 2.89 percentage points, and raise five-shot GSM8K accuracy from 23.96 to 36.16 percent. Evaluations on models from 1B to 70B parameters and on held-out C4 test the method beyond this setting. The evaluated FP32 implementation adds 2.25 percent to 4K prefill time; a separate compact-storage implementation retains a 27.7 percent peak-memory reduction relative to FP16.